All trend templates

Trendeo is a growing library of template-specific AI video generators. Each page packages the source motion, camera rules, cast positions, photo requirements, and prompt for one trend. Pick the format that fits your cast, then choose an eight- or fifteen-second output in 480p or 720p.

Live now

Two generators, one casting workflow.

Orange Booth Duet is a prepared two-person performance with source audio. Daylight Carpool is Trendeo’s original in-car scene for two to four riders. They share the same credit balance and privacy rules.

Two performers at a microphone in an orange recording booth

Orange Booth Duet

Two photos enter a prepared microphone-duet performance. Keep the timing, source sound and close two-person framing.

Open generator

Daylight Carpool

Put two to four people into a vertical carpool performance, with each uploaded photo assigned to a fixed seat.

Open generator

Beyond ready-made templates

Bring your own performance.

AI Video Recast replaces people in your own 5–30 second MP4 using reference photos, while aiming to preserve its motion, cuts and source audio. It is a separate H3 Max workflow, available at 768p and 1080p.

Open Recast ↗

Choose the right template

Match the clip to the people you have.

Use Hotel Lobby for a duo

Choose Hotel Lobby when you have two friends, a couple, two brothers, a pet and owner, or one person you want to duplicate across the performance. The camera stays locked and the left and right roles remain distinct.

The Hotel Lobby guideincludes six copyable prompt variants, photo advice, the Genjutsu method, and fixes for face swapping.

Use Walk It Talk It for a group

Choose Walk It Talk Itwhen seat placement is part of the joke. Start with two people in front or fill all four seats. The prompt pins each photo to a seat and asks the model to keep the street outside coherent.

Both templates need clear front-facing JPG or PNG photos under 3 MB. You never need to upload the original source video; that is already attached to the template.

What gets a page

A repeatable performance, not every passing meme.

Recognizable motion

The performance has a stable camera, clear roles, and a short source clip that survives identity replacement.

A useful cast map

Every visible person can be described as left, right, front, or back so the model receives an unambiguous job.

A real reason to return

New templates join the index when they offer a distinct format, not when they only repeat the same generator under another keyword.